.jpg)
It’s been a bit since our last update! In the interim we’ve welcomed two new artists to the program, as well as lined up four more to start over the coming months - things have been busy! We are very excited to have Anne Horel and Inès Kivimäki on board. More on them below!
Over the past few months we’ve been in talks with a16z’s Cultural Leadership Fund to help develop new programming to support Black artists. Through their guidance and connections, we are proud to announce our new partnership with the Black Digital Arts Collective (BDAC), who will be sponsoring a cohort of up to 8 artists in our AiR program through the end of this year! We are incredibly grateful to be partnered with such a talented and visionary people, and are looking forward to our first BDAC portfolio artist to start in August - Keion Kopper! You can find more on Keion here, and of course, we will be covering his work in the program in this newsletter come August.
Also, through our partnership with BDAC, we are incredibly excited to announce that will be working towards a Civitai AiR presence at Art Basel Miami this year! The fair takes place December 6-8, and an exhibition of work by AiR Artists will be on view during that time frame. More info on that in the months to come.
We're excited to welcome Anne Horel to as one of our new AiR Artists! Anne is a Paris-based multimedia artist known for her innovative fusion of digital, AI, and physical art forms. Her practice encompasses a wide range of media, including video, installation, and AI-generated imagery. With a background in both commercial work and contemporary art, Anne seamlessly blends elements of pop culture, fashion, and technology to create works that are both thought-provoking and visually captivating. Her playful yet poignant approach often explores themes of identity, consumerism, and the relationship between the virtual and the real. Follow her work at annehorel.com and on Instagram @annehorel.
Anne's residency proposal involves transforming a previously created body of AI works - masks, you might call them - into a series of physical plush toys! This whimsical project appealed to us from both technical and conceptual angles, as it aligns with Anne's ongoing exploration of consumerism, fashion, and design trends while also bridging the digital/physical divide in a new and interesting way. Our plan is to release these plush toys for sale on the civitai.com shop in a limited edition of 25 per design, with 5 different designs available to collect.
She has already built out a series of potential collections on her Figma board!
In our first official studio visit Anne began by showing me a collection of images she had created using an early beta version of DALL-E, one that is no longer publicly available. "These denim-textured creatures are simpler," she explained, pulling up a series of images that looked like they had been stitched together from worn jeans. "I wanted to start with something manageable, something that wouldn't overwhelm the viewer right away." We discussed denim, a material with a rich cultural history, as a symbol of transformation and reuse, weaving narratives of sustainability and fashion into her work.
We spent some time discussing the practicalities of bringing these AI-generated images into the physical world. She mused about using recycled fabrics to create tangible versions of these characters. "Imagine walking into a thrift shop in some remote part of the world and finding the perfect pair of jeans for a sculpture, it’s like the creatures themselves are living a second life through this material." We are looking into sourcing the denim for her plushy sculptures to give to our production partners, which would give each piece a unique vintage patina of worn denim.
She also presented more whimsical creations—anthropomorphic food items with a surreal twist. Anne laughed as she described them, "They’re wacky and funny, but there's an artsy side to them too." She pulled up a digital image of a rutabaga with legs and a hat, its expression somewhere between comical and oddly dignified. "These pieces are about finding humor in the everyday, turning mundane objects into characters with stories of their own."
Anne’s process is highly visual and intuitive, "these creatures don’t come from a story, but they tell one once they exist," she said. Her approach is almost reverse-engineered—creating the image first and then weaving a narrative around it. This method allows her to explore a wide range of themes, from fast fashion to environmental consciousness. She emphasized that the narratives in her work often emerge after or during the process of creation. In a moment of irony, we discussed her veganism and how these new works almost perform as hunting trophies culled from a digital landscape.
As we wrapped up our first visit, Anne shared her vision of collaborating with a fashion designer to bring her AI-generated creations to the runway. "I imagine these creatures as wearable art - a denim dress or an outfit designed by Jeremy Scott featuring these motifs—it would be incredible."
We are thrilled to welcome Inès Kivimäki to the Civitai AiR residency program!
Inès Kivimäki, born in 1990 in France, is a talented Finnish-Algerian artist whose dynamic practice spans a wide range of mediums and themes. Currently based in Los Angeles, Inès recently completed her MFA at UC Riverside, building on her BFA from Art Center College of Design in Pasadena. You can find examples of her recent work on her website.
Inès’s work is deeply rooted in thorough research and extends into various forms, reflecting her keen interest in the intersection of emerging and ancient technologies, device dependency, and its profound implications on self-image. Her art practice navigates these complex themes, offering insightful and often provocative explorations that challenge our perceptions of technology and identity.
In our AiR program, Inès will develop a project that uses a looped chain of AI models to conjure an entity or impossible perspective, inspired by magical thought and folklore. This innovative project will employ AI vision to analyze and reinterpret found images of simple objects from the Slavic folk tale of Baba Yaga - a hairbrush, an egg, a needle, a ring, butter - transforming them through different stages of text, image, and sound, then back again in a game of technological telephone culminating in a video loop output. Through this work, she will explore the nuances and errors in AI outputs, mirroring human cognition and challenging the boundaries of empirical thinking.
Inés is particularly fascinated by the way these myths have been transmitted through oral traditions and later documented in texts across Russian, Ukrainian, Finnish, and Karelian cultures. This project seeks to parallel AI technology with magical thinking, drawing connections between AI's "black box" nature and the mysterious, transformative processes of folklore as cultural production.
Jason has made significant progress over his 6 weeks in the program. So far he has built a working relationship with a local installation of A1111 (no small task), trained several of his own image models on Civitai, used A1111 to generate new compositional materials, composited those materials into two finished painting mockups, and has just this week begun his first large scale painting based on images fully generated through AI toolings. His first two painting mock ups areshown in evolutionary steps on his Figma board below:
To build these mock ups, Jason used Photoshop to remove backgrounds and composite the layers. Currently each painting has up to 16 layers of generated imagery. Jason emphasized the importance of balancing the AI-generated elements with traditional painting techniques to maintain a cohesive and compelling composition. His approach is highly iterative, and he's done an excellent job documenting each step in Figma to provide a transparent look at his creative process.
One of the newest elements emerging from this new precess is the sticker-like Cheshire Cat from "Alice in Wonderland," which he generated using a model from Civitai. Floating like a mischievous ghost over his composition, it confronts the viewer with a superficial grin: a doorman watchful of our entry into the visual field of the work. For me, this gesture extends the psychological space of the work and forces a sort of reflexivity into the mind of the viewer. In our gazing state, we are aware that we’re being watched.
Finally, here is a quick snapshot of the underpainting of his newest work. Very intrigued to see how it unfolds!
Drew began his residency by training a model on the same image 100 times, which immediately broke our moderation system and was taken down :). We were able to quickly restore the model and his experiments have yielded interesting results. The original image was a 4x5 film photograph taken out his studio window, focused on an 8-bit Duck Hunt decal he placed there. With very simple prompting, the on site generator produced a bevy of subtle variation, which he layered on top of one another to make a short gif, which you can check out on his Figma board. The first results are here:
Drew and I have engaged in some deep theoretical discussions this past month, exploring the intersections of his work with photography and AI. One of the recurring themes in our discussions was the concept of the AI as a co-creator, a sort of agential being that acts upon us as much as we act upon it.
Drew shared his view that working with AI is a two-way street: while the artist trains the AI with data and prompts, the AI, in turn, "trains" the artist by presenting a new "update" to our frame of reference that inescapably merges into our present reality. As Drew put it, "I make a thing or I encounter a technology that teaches me something about the world. And then I now live in a world that is inescapably infused with that information. And then I make a new thing or I encounter a new thing. And so it's cyclical."
Drew is exploring this phenomenon by training a model on a particular photographic diptych (that he created using a drone as a partner) which features the shape of a window arch seen from the inside and outside. This potentially offers new ways to think about composition, form, and abstraction. Drew explained, "The window shape is so prominent, if you get the filtration just right, it has a lot of leeway to do random stuff based on what your description is, but it maintains that shape."
The implications of this collaborative process have a deep impact on traditional notions of authorship. Drew posits that AI could be seen as an extension of the artist's mind, amplifying their ability to generate and manipulate visual information. This raises questions about the nature of creativity itself. Is creativity an exclusively human trait, or can it be shared with machines? Drew's work suggests that the answer might lie somewhere in between, where AI becomes a new kind of medium through which artistic thought can flow.
Photography, with its rich history of technological innovation, provides a fertile backdrop for these discussions. Photography has always been at the forefront of integrating new technologies, from the invention of the camera to digital imaging and now AI. Drew’s AI experiments are a natural extension of this lineage, pushing the medium into uncharted territories. We discussed the idea that just as photography transformed our perception of reality, AI-generated art has the potential to redefine our understanding of what is possible in visual art. Drew noted, "I think about that a lot. That's actually a version of a thing I talk about when I give lectures. When we invented the synthesizer, we were now introduced to a whole new world of sound that we didn't know about before."
Because of the highly repetitive nature of his subject matter, we have discussed the similarity of his outputs to film frames. He suggested that he might continue his experiments in the gif format to capture some of what I’ll call “programmatic abstraction” at play in his images.
We are also planning a conversation with Drew in the coming weeks on twitch, so stay tuned for that!
Ira has continued working on a few paintings which he likens to the process of diffusion. Ironically, these paintings make no use of AI in their technical workflow but instead stand in as a conceptual representation of what AI does by pulling details out of a latent space. You can see this process of emergence happening as he lays down thick brush strokes and then continues to refine, shifting the pallet slightly from day to day. In fact, in one of our visits we discussed how the shifting of the colors almost embodies the varying hues of daylight that would play across this imaginary subject from day to day, were it real.
This process of abstraction into figuration is somewhat unique in that the work could be seen as the process of creating the painting rather than the final result. At the end what we receive is a painting embedded with multiple daily permutations (meditations?) painted into it over and over again as a palimpsest of sorts, emblematic of the iterative process that Ira engages with daily in AI. I love this physical representation of his digital process.
He’s also been working on a new series of self portraits, where he prompts the AI to deliver him “the portrait of an artist named Ira Greenberg” (which I think would be a solid didactic title for the series). He then manipulates this image heavily through a series of img2img modifications and specific prompting to arrive at a shattered vision of “himself”, occluded by glassy fragments that often interrupt the subject’s features and therefore our ability to fully identify with the person we’re seeing.
I find them quite compelling, specifically because he envisions printing these very large - think Chuck Close scale. By coopting its language, the blurred fragments that obscure the face (particularly when they cover the eye) speak to AI's shattering of certain segments of contemporary photography. The defocused reflections act as a screen to intercept the viewer's gaze while moving the works solidly into the realm of abstraction. I can imagine the impact and immediacy of these works would have if printed at 60" x 60" and mounted on floating dibond with no frame. If anyone has a physical space to show these in, we are searching for a gallery context.
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)
.jpg)